Genetic analysis of individual origins supports isolation of grizzly
bears in the Greater Yellowstone Ecosystem
Mark A. Haroldson1,7, Charles C. Schwartz1, Katherine C. Kendall2, Kerry A. Gunther3,
David S. Moody4, Kevin Frey5, and David Paetkau6
1
United States Geological Survey, Northern Rocky Mountain Science Center, Interagency Grizzly Bear Team,
2327 University Way, Suite 2, Bozeman, MT 59715, USA
2
United States Geological Survey, Northern Rocky Mountain Science Center, Glacier Field Station,
Glacier National Park, West Glacier, MT 59936, USA
3
Bear Management Office, Yellowstone Center for Resources, PO Box 168, Yellowstone National Park,
WY 82190, USA
4
Wyoming Game and Fish Department, 260 Buena Vista, Lander, WY 82520, USA
5
Montana Fish, Wildlife and Parks, 1400 South 19th Avenue, Bozeman, MT 59718, USA
6
Wildlife Genetics International, Box 274, Nelson, British Columbia, V1L 5P9, Canada
Abstract: The Greater Yellowstone Ecosystem (GYE) supports the southernmost of the 2
largest remaining grizzly bear (Ursus arctos) populations in the contiguous United States. Since
the mid-1980s, this population has increased in numbers and expanded in range. However,
concerns for its long-term genetic health remain because of its presumed continued isolation. To
test the power of genetic methods for detecting immigrants, we generated 16-locus microsatellite
genotypes for 424 individual grizzly bears sampled in the GYE during 1983–2007. Genotyping
success was high (90%) and varied by sample type, with poorest success (40%) for hair collected
from mortalities found .1 day after death. Years of storage did not affect genotyping success.
Observed heterozygosity was 0.60, with a mean of 5.2 alleles/marker. We used factorial
correspondence analysis (Program GENETIX) and Bayesian clustering (Program STRUCTURE) to compare 424 GYE genotypes with 601 existing genotypes from grizzly bears sampled
in the Northern Continental Divide Ecosystem (NCDE) (FST 5 0.096 between GYE and
NCDE). These methods correctly classified all sampled individuals to their population of origin,
providing no evidence of natural movement between the GYE and NCDE. Analysis of 500
simulated first-generation crosses suggested that over 95% of such bears would also be
detectable using our 16-locus data set. Our approach provides a practical method for detecting
immigration in the GYE grizzly population. We discuss estimates for the proportion of the
GYE population sampled and prospects for natural immigration into the GYE.
Key words: Bayesian clustering, DNA, factorial correspondence, grizzly bear, immigration, Northern
Continental Divide, Yellowstone, Ursus arctos
Ursus 21(1):1–13 (2010)
The abundance and range of grizzly bears (Ursus
arctos) in the contiguous 48 states has declined
during the 20th century. Grizzly bears were eliminated from 98% of their historic range (Mattson et
al. 1995), and 31 of 37 bear populations recognized
in 1922 were eliminated by 1975 (Servheen 1999).
For grizzly bears in the Greater Yellowstone
Ecosystem (GYE), the effect was one of increasing
isolation. By 1959, when Craighead et al. (1995)
began their pioneering work on grizzly bears in
Yellowstone, the population had been reduced to a
fraction of its former size and was relegated largely
to Yellowstone National Park (YNP) and surrounding environs. High grizzly bear mortality in 1970 and
1971 following closure of open-pit dumps in YNP
(National Academy of Sciences 1974) combined with
uncertainty about population status prompted the
US Fish and Wildlife Service (USFWS) in 1975 to
list the species as threatened south of Canada under
the Endangered Species Act (USFWS 1993). After
listing, the population continued to decline (Knight
and Eberhardt 1984, 1985, 1987) until strategies
7
[email protected]
1
2
DNA SUPPORTS ISOLATION OF GYE GRIZZLY N Haroldson et al.
designed to reduce human-caused mortality were
implemented in the mid-1980s (Interagency Grizzly
Bear Committee 1986).
Evidence from 2 independent datasets indicates that
grizzly bear numbers in the GYE have increased since
the mid 1980s. Counts of unique females with cubs-ofthe-year (unduplicated females as per Knight et al.
1995) have increased (Harris et al. 2007), and
consistent with this trend, estimates of l derived from
radio-monitored bears also indicate a positive population trend (Eberhardt et al. 1994, Eberhardt 1995,
Boyce et al. 2001, Harris et al. 2006). Concurrent with
this population increase, bears have continued to
expand their range (Blanchard et al. 1992; Schwartz et
al. 2002, 2006a; Pyare et al. 2004).
However, the population remains isolated and
concerns for its genetic health continue. Miller and
Waits (2003) reported a slight decline in genetic
diversity for Yellowstone bears since the early 20th
century. They concluded that, given current population size, it was unlikely that genetic factors would
negatively impact population viability in the near
term. In the long term, they suggested that 1–2
migrants/generation from the Northern Continental
Divide Ecosystem (NCDE) of northwest Montana
(the geographically closest population) would maintain diversity.
Here we report on results of our initial efforts to
determine if natural immigration into the GYE has
occurred in recent years. We employed standard
genetic techniques to analyze 16 microsatellite
markers for a large number of grizzly bear samples
obtained in the GYE. We used assignment tests and
compared population-specific genotypes (Eldridge et
al. 2001) from the GYE to those from the NCDE
(Kendall et al. 2008, 2009) to identify the probable
population of origin for individual bears. We also
estimate our ability to detect immigrants in future
monitoring based on our sampling intensity during
1983–2007.
Study area
We considered the landscape supporting grizzly
bear populations in the GYE and the NCDE our
study areas (Fig. 1). The GYE encompassed Yellowstone and Grand Teton National Parks (GTNP) plus
portions of 6 adjacent national forests (Beaverhead–
Deerlodge, Bridger–Teton, Caribou–Targhee, Custer, Gallatin, and Shoshone) and smaller amounts of
state and private land in Wyoming, Montana, and
Idaho, USA. Geographically, the GYE included
headwaters of the Missouri–Mississippi, Snake–
Columbia, and Green–Colorado river systems, the
Yellowstone Plateau, and 14 surrounding mountain
ranges (Marston and Anderson 1991). Detailed
descriptions can be found in Blanchard and Knight
(1991), Mattson et al. (1991), and Schwartz et al.
(2006b).
The NCDE, located in the northern Rocky
Mountains of Montana, included Glacier National
Park, portions of 5 national forests (Flathead,
Kootenai, Lewis and Clark, Lolo, and Helena),
and 2 Indian Reservations (Blackfeet Nations, and
Confederated Salish and Kootenai Tribes). Geographically, the NCDE encompassed 9 mountain
ranges and included watersheds of the Flathead–
Columbia, Missouri–Mississippi, and St. Mary–
Saskatchewan river systems. Climates and associated
vegetation varied, with distinctive Pacific maritime
influences in the northwest and continental influences in the southeast (Daubenmire 1969). Climatically distinct portions of NCDE with associated key
grizzly bear foods were described in Mace and
Jonkel (1986).
During 1990–2004, grizzly bears in the GYE
occupied roughly 37,000 km2, and evidence suggested
continued range expansion (Schwartz et al. 2006b).
Occasional occurrences, usually in the form of
conflicts or mortalities, were documented well beyond
the area considered occupied (Interagency Grizzly
Bear Study Team [IGBST] annual reports at http://
nrmsc.usgs.gov/products/IGBST). Kendall et al.
(2009) concluded that grizzly bears in the NCDE were
also expanding their range and estimated the population occupied 33,480 km2 during 1994–2007. Using
these most recent estimates of occupied range
(Schwartz et al. 2006b, Kendall et al. 2009) suggests
that a straight-line distance of approximately 165 km
separated the GYE from the NCDE population to the
north (Fig. 1). Several major transportation corridors,
including the east–west Interstate Highway 90 and the
north–south Interstate Highway 15, occurred in the
intervening terrain (Fig. 1). These routes generally
follow broad valley bottoms in which livestock
husbandry was the predominant activity.
Methods
GYE samples
Samples for genotyping were collected by bear
researchers and managers from live captures and
Ursus 21(1):1–13 (2010)
DNA SUPPORTS ISOLATION OF GYE GRIZZLY N Haroldson et al.
3
Fig. 1. The Greater Yellowstone (GYE) and Northern Continental Divide Ecosystems (NCDE), Idaho, Montana,
and Wyoming, USA. Shaded areas are estimated current range for each grizzly bear population. Average
locations for individual genotypes were estimated from geographic means of VHF telemetry locations (when
available) in the GYE, and geographic means of location for hair collections for individuals in the NCDE. Gray
lines show interstate highways.
mortalities throughout the GYE. Capture methods
typically employed culvert traps or Aldrich leg-hold
snares and are described in Schwartz et al. (2006b). A
small number of hair samples (n 5 18) were obtained
using hair corrals where bears were not handled. These
samples were obtained from an investigation of a
suspected adoption by a female grizzly with cubs in
YNP during 2007 (Haroldson et al. 2008a).
Type of sample materials varied over the duration
of study but generally included hair, tissue, or blood.
We placed hair samples in manila envelopes during
handling and stored them in file cabinets affixed to
original field forms at room temperature. We froze
ear plugs at 232uC. We stored blood samples
collected prior to 1996 as clots with ethylene glycol
at 232uC. After 1997, we stored clots and some
blood samples in Longmire lysis buffer solution
(Longmire et al. 1988) at room temperature. The
Wyoming Game and Fish Wildlife Forensic and Fish
Ursus 21(1):1–13 (2010)
Health Laboratory also provided DNA material
(n 5 53) extracted from samples using standard
phenol:chloroform and Eppendorf PhaseLock tubes
or Super Quick Gene (ATCG, Fort Collins, Colorado, USA). They stored extracted samples in TE
(Tris buffer + Ethylene diamine tetraacetic acid)
buffer at 270uC (D. Hawk, Wyoming Game and
Fish Department, Laramie, Wyoming, USA, personal communication, 2008).
Some samples were collected before genetic
methods were widely available. Thus, no special
consideration for genetic methodology was employed when samples, particularly hair samples,
were originally archived. We used this as an
opportunity to investigate the effect of storage time
(in years) on genotyping success of hair samples. We
used binary logistic regression (MiniTab version
14.20) to test genotyping success as a function of
storage time in years.
4
DNA SUPPORTS ISOLATION OF GYE GRIZZLY N Haroldson et al.
Genetic methods. We used material from the
most recent handling for known individuals with .1
capture (i.e., marked). After the initial sample
selection, we made a random draw of 10 samples
without replacement from individuals with multiple
samples as blind tests for individual identification.
We extracted DNA using DNeasy Tissue kits
following manufacturer’s instructions (Qiagen Inc.,
Mississauga, Ontario, Canada, http://www.qiagen.
com/) at a laboratory (Wildlife Genetics International, Nelson, British Columbia, Canada) specializing in the analysis of low quantity and quality
DNA material (Woods et al. 1999, Paetkau 2003,
Roon et al. 2005). We genotyped successful DNA
extractions for individuals at 16 well documented
microsatellite sites (Ostrander et al. 1993, Taberlet et
al. 1997, Paetkau et al. 1998, Proctor et al. 2002)
used by Kendall et al. (2008, 2009): G1A, G10B,
G10C, G1D, G10H, G10J, G10L, G10M, G10P,
G10U, G10X, MU23, MU50, MU59, CXX20, and
CXX110. Genotyping was completed by analyzing
markers in 3 groups. Samples were considered
unacceptably error-prone and culled from subsequent analysis if they failed at .5 of 11 markers
during the first round of genotyping (Paetkau 2003).
After the first group of markers was analyzed,
individual identification was established based on a
minimum of 9 markers for which all samples had
complete genotypes. We prevented the identification
of false individuals (genotyping error) by re-analyzing mismatching markers for pairs of genotypes that
matched at 7 or 8 of 9 markers, an approach that
was validated through extensive blind testing during
the NCDE analysis (Kendall et al. 2009). Only 1
sample per individual had its genotype extended to
the full set of 16 markers.
NCDE samples
Samples from the NCDE to investigate immigration into the GYE were from grizzly bear hair
collected and analyzed as part of the Greater Glacier
DNA study (Kendall et al. 2008) or the NCDE DNA
effort (Kendall et al. 2009). Combined results from
these studies produced 601 individuals with complete
genotypes at the 16 microsatellite markers. During
September 2005, a subadult male grizzly bear was
found dead in Mill Creek, southwest of Anaconda,
Montana, approximately 80 km south of the
estimated distribution for NCDE bears, and 120 km
northwest of the distribution of GYE bears.
Montana Fish, Wildlife and Parks provided a
15-locus genotype for this bear that we included in
our analysis.
Statistical analysis
We estimated variability within populations,
differentiation between populations, and the significance of departures from Hardy-Weinberg equilibrium using Genepop 4.0.10 as implemented at
‘‘Genepop on the Web’’ (Rousset 2008; http://
genepop.curtin.edu.au/). We used a multidimensional factorial correspondence analysis (FCA) within
Program GENETIX (Belkhir 1999) to identify
clusters of individuals with similar genotypes. The
procedure provides an objective method for determining the birth population by making no a priori
assumption of group membership. Individuals are
grouped on multiple factorial axes based on shared
alleles. We also used the Bayesian clustering
Program STRUCTURE (Pritchard et al. 2000) to
look for evidence of genetic exchange between the 2
populations. Program STRUCTURE uses a Markov
chain Monte Carlo (MCMC) process to subdivide
data into likely populations, then estimates the
proportion of each individual’s ancestry from each
population. We used the admixture model with 1
million MCMC steps (including a 250,000-step burnin period) and with k (the putative number of
populations) set to 2. STRUCTURE is often used to
estimate the number of populations in a dataset by
exploring a range of values of k, but because our
goal was to separate GYE bears from non-GYE
bears, we did not explore higher values of k.
We also tested the power of STRUCTURE to
detect first-generation hybrids by simulating 500
hybrid genotypes by drawing 1 allele at random from
the allele distributions observed in the GYE and
NCDE datasets using a random number function in
Microsoft Excel. This process was repeated for each
marker to create 16-locus genotypes with 1 allele
from each population at each locus. Drawing alleles
at random should provide a reasonable reflection of
detection power in a system with no recent history of
immigration, but if immigrants enter the Yellowstone population, it will become necessary to employ
more sophisticated sampling methods that account
for gametic disequilibrium (Paetkau et al. 2004).
Hybrid genotypes were analyzed by adding them
in sets of 10 to the original STRUCTURE input file
containing data from the 424 + 601 actual bears,
based on the observation that larger additions of
hybrid genotypes reduced the power of the program
Ursus 21(1):1–13 (2010)
DNA SUPPORTS ISOLATION OF GYE GRIZZLY N Haroldson et al.
to apportion the ancestry of non-hybrid (real)
individuals.
Probability of detection. We estimated the
cumulative probability of detecting a new migrant
to the GYE population over time using a given
amount of search effort. First, we estimated the
proportion of the population genotyped in 2007. We
derived this proportion by summing new captures,
mortalities, and bears known to be alive because of
radiotelemetry in 2007. These 122 samples represented 21% of the total standing population estimated that year (571; Haroldson 2008).
We further estimated the number of new individuals that would become available for genotyping
after 2007 if we continue to sample at our current
rate. For 1998–2007, the IGBST, in conjunction with
state wildlife management agencies (Idaho, Montana, Wyoming) and National Park Service managers (GTNP, YNP), annually trapped and handled 36
new individuals (IGBST annual reports, http://
nrmsc.usgs.gov/products/IGBST) and documented
an average of 9 mortalities from unmarked individuals. Thus, on average, we sampled approximately
45 new independent bears (.2 years of age) each
year. We assumed the population would continue to
grow at current rates (estimated point l 5 1.04,
Harris et al. 2006) and that survival for independent
bears were similar to estimates (male 5 0.874, female
5 0.95) by Haroldson et al. (2006). We also assumed
we would only sample independent bears because the
IGBST rarely captured dependent young. We used
Program RISKMAN (Taylor et al. 2001) with
demographic parameters for the GYE population
to estimate the proportion of dependent young
(0.249), independent males (0.294), and independent
females (0.458). In 2007, we estimated there were 142
dependent young (571 x 0.249) and 429 independent
bears (male 5 168, and female 5 261). Therefore, we
estimated there were 307 independent bears that had
not been genotyped (429 - 122). Thus, we estimated
the probability of detecting 1 new migrant in 2007 as
0.146 (45/307). Carrying these calculations forward,
we applied the estimated l to the estimated
population size and reduced the number of known
marked bears by the weighted mortality rate for
independent male and female bears in the population. We then estimated the probability of detecting
a new individual in 2008 and subsequent years. We
estimated the annual cumulative probability of
detecting a migrant entering the population in 2007
using the formula:
Ursus 21(1):1–13 (2010)
5
n
1{( P (1{Xt ))
t~1
where t 5 year, beginning in 2007, and
Xt 5 the annual probability of detecting a new
individual (adapted from McArdle 1990).
Results
GYE genotyping
We successfully genotyped 451 of 496 samples.
Accounting for samples with identical multilocus
genotypes (assumed to come from the same individual), our final dataset contained complete 16-locus
genotypes for 424 unique individuals. Genotyping
success varied by sample type (Table 1), with hair
samples having the lowest success rate (90%) and
whole blood or tissue samples having the highest
(100%). Among different sources of hair, the largest
percent of failures (40%) were from dead bears when
samples were obtained .1 day after death (Table 1).
Samples were obtained from 1983–2007, with the
largest number (n 5 86, 17%) coming from 2007
because we typically used the most recent sample
available from each bear. Among hair samples from
captured bears with adequate material (Table 1, n 5
294), there was no evidence that the number of years
samples were stored (range 5 0–24 yrs) affected
genotyping success (G 5 2.532, 1 df, P 5 0.112).
This suggested that other factors, possibly condition
of samples at the time of archiving, had a greater
effect on genotyping success than storage time.
We included 10 samples (2%) from randomly
chosen individuals who had been captured more
than once as blind tests for individual identification.
Of these, 9 were correctly identified as duplicated
samples and 1 pair failed due to poor quality. We
also successfully resolved the identities of 4 individuals with lost or inconclusive marks (lost tags or
unreadable tattoos) at handling. In each instance,
differences in estimated ages (from cementum
annuli count of sectioned teeth, Stoneberg and
Jonkel 1966) between initial capture and subsequent
capture were consistent with the genetic results
identifying the samples as having come from a
single individual. For example, we captured bear
204 at age 2 in 1992 and a 17-year-old bear 15 years
later in 2007 with no marks (designated 575).
Genotyping revealed that 575 and 204 were
identical; their age difference was consistent with
the number of years between sampling.
6
DNA SUPPORTS ISOLATION OF GYE GRIZZLY N Haroldson et al.
Table 1. Grizzly bear samples analyzed and successfully genotyped by sample origin and type, Greater
Yellowstone Ecosystem, 1983–2007.
Genotyping result
Sample origin
Sample type
Total
Sampled at capture
hair
blood (buffered)
blood (EDTA)c
blood (frozen clot)
tissue (ear plug)
extracted DNA
298
53
2
46
4
53
Hair snag
hair
Found mortality sampled
,1 day after death
hair
tissue (muscle)
Found mortality sampled
.1 day after death
Total
hair
20
9
1
10
496
Successful (%)
268
53
2
45
4
49
(89.9)
(100.0)
(100.0)
(97.8)
(100.0)
(92.5)
17 (85.0)
8 (88.9)
1 (100.0)
4 (40.0)
451 (90.9)
Failed Inadequatea
Mixedb
Black bear
26
0
0
1
0
4
3
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
0
1
1
1
0
0
0
0
0
0
0
5
38
1
4
0
2
0
1
a
Sample lacked suitable material for extraction.
Sample contained material from multiple bears.
c
Blood preserved in ethylene diamine tetraacetic acid (EDTA).
b
Forty-six samples were replicated because they
represented 2–8 samples from 19 individuals. These
samples yielded a total of 628 single-locus genotypes
(not all samples were genotyped at all 16 markers)
for comparison. We detected a single error, giving an
estimated error rate of 0.0016 (1/628). This error
rate/sample/locus multiplied by the number of
genotypes that were not replicated as positive
controls (,390) suggested approximately 10 errors
in the dataset. Thus, we expect that roughly 1 in 40
individuals might have contained an error at 1 of the
16 markers. This rate of error is not expected to have
meaningful effects on analysis of individual origins.
We expected the error rate in the 601 genotypes from
the NCDE to be lower yet because those 16-locus
genotypes were run in duplicate (Kendall et al.
2009). Kendall et al. (2009) reported an error rate of
0.002 for all loci/sample.
Because most of the GYE samples genotyped were
from marked (i.e., handled) individuals, the chance
of assigning samples to the same individual when
they were in fact different bears was small. Still, we
evaluated this potential for false matches using the
distribution of genotype similarity (‘‘mismatch distribution’’, i.e., 1MM-pairs differ at 1 marker,
2MM-pairs differ at 2 markers, Paetkau 2003). We
checked for matching genotypes before running all
16 markers, so we conservatively based our mismatch distribution on only the 9 markers for which
every sample was analyzed. Our graphed mismatch
distribution (Fig. 2) indicated 12 2MM-pairs
(matches at 7 of 9 markers) among 424 individuals.
By extrapolating the curve, we would not expect to
observe any 1 MM-pairs, and this was true — there
were no matches at 8 of 9 markers. Matches at all 9
markers would be even less common, suggesting
little risk that we sampled 2 individual bears with the
same 9-locus genotype.
Fig. 2. Mismatch distribution for 424 grizzly bears
in the Greater Yellowstone Ecosystem, 1983–2007,
based on 9 markers that were used for establishing
individual identities. Extrapolation suggests no
realistic chance of different bears having the same
genotype at 9 or more loci.
Ursus 21(1):1–13 (2010)
DNA SUPPORTS ISOLATION OF GYE GRIZZLY N Haroldson et al.
Table 2. Marker variability expressed as expected
(H E ) and observed (H O ) heterozygosity, and
observed number of alleles (A) for 16 microsatellite
loci from 424 individual grizzly bears from the
Greater Yellowstone Ecosystem, 1983–2007.
Locus
G1A
G1D
G10B
G10C
G10H
G10J
G10L
G10M
G10P
G10U
G10X
MU50
MU59
CXX20
CXX110
MU23
Mean
HE
HO
A
0.66
0.80
0.72
0.46
0.57
0.64
0.40
0.65
0.74
0.63
0.08
0.60
0.75
0.52
0.57
0.81
0.60
0.63
0.79
0.75
0.44
0.60
0.64
0.44
0.64
0.76
0.65
0.08
0.61
0.75
0.48
0.57
0.81
0.60
6
7
6
5
6
4
2
6
5
5
2
6
7
3
6
7
5.2
Genetic diversity and divergence
Our estimates of observed and expected heterozygosity using 16 markers for 424 genotyped bears
were 0.60 and 0.60 (NCDE 5 0.67 and 0.68),
respectively (Table 2). We tested for heterozygote
deficit at each marker, with one marker returning P
5 0.03. One result with P , 0.05 in 16 tests is
expected given our nominal Type I error rate,
although a small heterozygote deficit would be
anticipated in the likely event that the study population was not perfectly panmictic across the
,37,000 km2 study area. Tests for linkage disequilibrium between each of 120 pairs of markers returned
48 with P , 0.05, which was inconsistent with our null
hypothesis. This appeared to support the (biologically
plausible) hypothesis that the GYE grizzly bear
population was not completely panmictic.
We found a mean of 5.2 alleles/marker (compared
with 7.5 in the NCDE). Marker G10X was
essentially fixed in GYE bears (H 5 0.08), but we
included it because it is useful in differentiating GYE
and NCDE populations. Estimated genetic distance
FST between the GYE and NCDE was 0.096, similar
to the highest values (0.09) described between
regions of the NCDE (Kendall et al. 2009).
Assignment tests
FCA analysis for 424 individuals from the GYE
and 601 NCDE bears produced non-overlapping
value ranges for the 2 populations on the first axis
Ursus 21(1):1–13 (2010)
7
(Fig. 3), providing strong evidence that no sampled
individuals had moved between these populations. In
addition, our FCA results (Fig. 3) indicated that the
bear found dead in Mill Creek during 2005 was
genetically similar to NCDE bears. Program
STRUCTURE also produced strong separation
between GYE and NCDE bears, with all NCDE
bears having ,21% of their estimated genetic
ancestry in the GYE, and all GYE bears having
.82% estimated ancestry in that population. Results
were stable between runs, with the estimated
proportion of any individual’s ancestry in the GYE
varying by ,0.004 across 3 runs of STRUCTURE.
Given the wide separation between GYE and NCDE
bears with STRUCTURE (no values between 0.21
and 0.82 GYE ancestry) compared to the near
overlap of clusters with the FCA analysis (Fig. 3),
we used STRUCTURE to explore the potential for
identifying F1 hybrids.
When the 500 simulated hybrid genotypes were
analyzed in Program STRUCTURE, only 21 had
estimated ancestry within the range observed for
actual GYE bears, suggesting that approximately
96% (1–21/500) of first-generation hybrid offspring
would be detectable by virtue of having values
outside the observed range of GYE bears.
Probability of detection
Our estimated cumulative probability suggested
that if a single migrant entered the GYE population,
our probability of detection given our sampling
intensity would have been about 0.147 in the year of
immigration, but by 5 years later that probability
would have increased to 0.589 (Fig. 4). The probability of detecting a first-generation hybrid during its
initial year of independence (age 2) was approximately 0.141 (0.147 x 0.96), and 0.565 after 5 more
years if it survived (Fig. 4).
Discussion
Overall, our genotyping success for GYE samples
was high (90%) and varied by sample type. The
poorest success (40%) was obtained using hair
collected from dead bears found .1 day after death.
This leads us to suggest that material other than hair
should be collected from dead bears for genotyping,
especially if the carcass shows any sign indicating
time of death was .1 day prior to discovery. Good
genotyping success using foot pads from dead bears
have been achieved (Wildlife Genetics International,
8
DNA SUPPORTS ISOLATION OF GYE GRIZZLY N Haroldson et al.
Fig. 3. Dimensions 1 and 2 Genetrix output comparing individual genotypes for grizzly bears from the
Greater Yellowstone (GYE) and Northern Continental Divide Ecosystems (NCDE).
Nelson, British Columbia, Canada, unpublished
data), and we suggest use of this tissue for bears
found .1 day after death. We found that the length
of time (in years) hair samples were stored was not a
factor in genotyping success. Condition of the hair
samples when initially stored likely had more impact
on later genotyping success than time stored.
Estimates for heterozygosity and allele frequencies
we obtained for GYE grizzly bears were similar to
previously published estimates (Paetkau et al. 1998,
Waits et al. 1998, Miller and Waits 2003).
We found no evidence of recent immigrants from
the NCDE or their first-generation progeny in GYE.
Our results provide clear evidence that the techniques we employed would have identified migrants
and most of their F1 offspring had we sampled them.
The detection of progeny would indicate functional
connectivity (immigration and breeding) between the
GYE and other grizzly bear populations. The use of
additional markers in the analysis or the passage of
time in the absence of gene flow (i.e., drift) would
both improve the ability to identify the genetic origin
Fig. 4. Cumulative probability of detecting a single new migrant or first-generation hybrid in the Greater
Yellowstone Ecosystem assuming 45 new individuals genotyped annually and assuming that the migrant
entered the population (or a hybrid juvenile .2 years old became an independent bear) in year 1.
Ursus 21(1):1–13 (2010)
DNA SUPPORTS ISOLATION OF GYE GRIZZLY N Haroldson et al.
of first-generation offspring. Similar methods using
15 microsatellite markers were used to determine
that a grizzly bear caught in the Canadian Flathead
Valley (NCDE) in 1996 was likely a first-generation
hybrid from NCDE and GYE parents (Wildlife
Genetics International, Nelson, British Columbia,
Canada, unpublished data). We know of 4 nuisance
bears, 1 of which was a female accompanied by 2
cubs, that were transplanted to the Canadian
Flathead from the GYE during 1982–83 (Knight
and Blanchard 1983, Knight et al. 1985). Fates for 2
of these bears after transport were unknown. It is
possible that 1 of these individuals survived and
produced the putative hybrid caught in the Canadian
Flathead. Additional support comes from captive
grizzly bears at Washington State University (WSU).
Fortin (2009) reported correct genetic identification
by the laboratory that we used of 1 NCDE bear and
2 offspring from a NCDE and GYE mating that
were included as blind tests among samples of YNP
grizzly bear hair analyzed for a separate study.
Given near certainty that we can identify an
immigrant or its progeny from the NCDE or other
populations (Proctor et al. in press) if sampled, the
probability of detection is a function of the
prevalence in the GYE of grizzlies originating in
other populations. Our cumulative estimates of
detection (Fig. 4) assume constant population
growth and mortality rates. If growth rate slows
and we continue to mark the same number of
individuals, our probability of detection would
increase unless mortality rates also increase and we
obtain fewer samples from bears that die. Our
cumulative estimate of detection also assumes equal
probability of capture among sex–age classes of
bears, which is unlikely. We know from trapping
records that subadult males are typically more
vulnerable to capture than other sex–age classes of
bears. We anticipate that a new migrant from the
NCDE would likely be a subadult male, making our
estimate of detection conservative. Finally, should a
new migrant breed with a GYE resident, the
probability of detecting a first-generation hybrid
would be high (0.96), but likely delayed by 3 years
until the offspring becomes independent because we
sample so few cubs and yearlings. We therefore
consider our sampling and genotyping effort sufficient to have a high probability of detecting natural
immigrants or their progeny.
We were not surprised that we did not detect any
immigrants from the NCDE. Studies involving
Ursus 21(1):1–13 (2010)
9
marked and radio instrumented grizzly bears have
been conducted almost continuously in the GYE
since 1959 (Craighead et al. 1995, Haroldson et al.
2008b) and since the mid-1970s in the NCDE (Jonkel
1982). In nearly 50 years of combined effort
involving hundreds of marked bears, natural movement between the GYE and other populations has
not been documented. This suggests that the distance
and the intervening terrain are significant barriers to
movement between these populations, even for a
highly vagile species such as the grizzly bear.
Natural connectivity between the NCDE and
GYE requires corridors or linkage zones which at
the minimum allow for some movement of individuals between these populations. Walker and Craighead (1997) used least-cost-path modeling to identify
3 potential corridors linking the NCDE grizzly
population to the GYE. These were through (1)
the Big Belt–Bridger–Gallatin mountain ranges, (2)
the Boulder–Tobacco Root–Gravelly–Taylor–Hilgard ranges, and (3) the Selway–Bitterroot–Lemhi–
Centennial–Madison ranges. Recently Cushman et
al. (2009) produced similar results, identifying
corridors 1 and 2 of Walker and Craighead (1997)
using least-cost-path modeling and a geneticallybased landscape resistance model for black bear
(Ursus americanus). Cushman et al. (2009) suggested
their black bear model was useful for predicting
movement corridors for grizzly bears because of the
similarity of their results with those of other studies
of grizzly bear linkage (Sandstrom 1996, Servheen et
al. 2001). Forest cover and road densities were
important themes in the development of resistance
layers for models of both Walker and Craighead
(1997) and Cushman et al. (2009). Thus it is not
surprising they identified similar potential linkage
routes. Cushman et al. (2009:373, Table 1) also
identified land ownership and potential barriers to
animal movements along their predicted routes.
Characteristics of grizzly bear dispersal are additional factors that will influence the potential for
natural immigration into the GYE. Grizzly bear
dispersal is usually sex-biased (Blanchard and
Knight 1991, McLellan and Hovey 2001, Proctor
et al. 2004, Zedrosser et al. 2007), with males
dispersing on average 3 times further than females
(McLellan and Hovey 2001, Proctor et al. 2004).
Mean male dispersal distances were 42 km and 70 km
in southwestern Canada and the GYE, respectively
(Blanchard and Knight 1991, Proctor et al. 2004),
and dispersal distances of up to 175 km have been
10
DNA SUPPORTS ISOLATION OF GYE GRIZZLY N Haroldson et al.
documented (Proctor et al. 2004). Male grizzly bears
are capable of more extreme movements. In the
Canadian Arctic a male bear was observed on Melville
Island, Northwest Territories, .600 km from the
mainland where the nearest female grizzly would likely
reside (Doupe et al. 2007). A result of this bias in
dispersal and movements is an asymmetric susceptibility to fragmentation between the sexes (Proctor et
al. 2004, 2005). However, the possibility of connectivity by at least male bears is the minimum requirement
to maintain the genetic health of the GYE population.
The ,165 km distance between occupied ranges for
these populations is 2–3 times the mean dispersal
distance, but approximately equal to the maximum
distances observed for male bears in the region.
Kendall et al. (2009) suggested that the NCDE
population expanded its range in recent years. If this
trend continues, suitable habitats south of the
current NCDE population along several possible
linkage zones may become reoccupied. Although
trends in range expansion in the GYE have been
more southerly (Pyare et al. 2004, Schwartz et al.
2006a, IGBST unpublished data), there is suitable
habitat to the north and west of their current range
that bears in the GYE continue to reoccupy. If the
distance between the NCDE and GYE populations
decreases, the potential for natural immigration to
the GYE increases. The most likely scenario for such
an event would involve a young male bear dispersing
from recently reoccupied habitat across 1 or more
transportation corridors. Supporting this hypothesis,
the subadult male bear found dead in September
2005 in Mill Creek southwest of Anaconda, Montana, was approximately 80 km south of the
estimated distribution for NCDE bears. This individual had NCDE ancestry; we believe it was
probably dispersing because the Pioneer Mountains
are not known to contain a resident population of
grizzly bears. In a similar result, Proctor et al. (in
press) reported that a grizzly bear killed by a black
bear hunter in the Bitterroot Mountains of Idaho
during the fall of 2007 likely originated in the Selkirk
Mountains of southern British Columbia and thus
had dispersed .240 km. The Bitterroot Mountains
are also not known to contain a resident grizzly bear
population.
Considering current trends in range expansion and
population increases, the potential for natural
immigration, at least by male bears, into the GYE
has likely increased. However, increases in human
development along potential linkage zones may
negate this increased potential unless secure travel
corridors are maintained among populations.
Acknowledgments
Funding and support for the Interagency Grizzly
Bear Study Team was provided by the US Geological
Survey, Northern Rocky Mountain Science Center;
US Fish and Wildlife Service; Wyoming Game and
Fish Department (WGFD); National Park Service;
Montana Department of Fish, Wildlife and Parks;
and the Idaho Fish and Game Department. Since
1997, grizzly bear capture and handling procedures
were reviewed and approved by the Animal Care and
Use Committee of the US Geological Survey,
Biological Resources Division; procedures conformed
to the Animal Welfare Act and to US Government
principles for the utilization and care of vertebrate
animals used in testing, research, and training.
Captures were conducted under USFWS Endangered
Species Permit [Section (i) C and D of the grizzly bear
4(d) rule, 50 CFR17.40 (b)] and YNP Research
Permit YELL-00073. We thank M. Bruscino,
WGFD, for providing grizzly bear samples. We also
recognize the dedicated field crews that safely handled
bears and procured the samples needed for this work.
We thank K. West for editing early drafts of the
manuscript and S. Schmitz for sample preparation.
Any use of trade, product, or firm names is for
descriptive purposes only and does not imply
endorsement by the US Government.
Literature cited
BELKHIR, K. 1999. GENETIX, v. 4.0. Laboratoire
Génome, Populations, Interactions. Centre National
de la Recherche Scientifique Unite Mixte de Recherche,
Université, Montpelier, France.
BLANCHARD, B., AND R. KNIGHT. 1991. Movements of
Yellowstone grizzly bears, 1975–87. Biological Conservation 58:41–67.
———, ———, AND D. MATTSON. 1992. Distribution of
Yellowstone grizzly bears during the 1980s. American
Midland Naturalist 128:332–338.
BOYCE, M.S., B.M. BLANCHARD, R.R. KNIGHT, AND C.
SERVHEEN. 2001. Population viability for grizzly bears:
A critical review. International Association of Bear
Research and Management Monograph Series 4.
CRAIGHEAD, J.J., J.S. SUMNER, AND J.A. MITCHELL. 1995.
The grizzly bears of Yellowstone: Their ecology in the
Yellowstone ecosystem, 1959–1992. Island Press, Washington, DC, USA.
Ursus 21(1):1–13 (2010)
DNA SUPPORTS ISOLATION OF GYE GRIZZLY N Haroldson et al.
CUSHMAN, S.A., K.S. MCKELVEY, AND M.K. SCHWARTZ.
2009. Use of empirically derived source-destination
models to map regional conservation corridors. Conservation Biology 23:368–376.
DAUBENMIRE, R. 1969. Structure and ecology of coniferous
forests of the northern Rocky Mountains. Pages 25–41
in R.D. Taber, editor. Coniferous forests on the
Northern Rocky Mountains. University of Montana,
Missoula, Montana, USA.
DOUPE, J.P., J.H. ENGLAND, M. FURZE, AND D. PAETKAU.
2007. Most northerly observation of a grizzly bear
(Ursus arctos) in Canada: Photographic and DNA
evidence from Melville Island, Northwest Territories.
Arctic 60:271–276.
EBERHARDT, L.L., B.M. BLANCHARD, AND R.R. KNIGHT.
1994. Population trend of the Yellowstone grizzly bear
as estimated from reproductive and survival rates.
Canadian Journal of Zoology 72:360–363.
———. 1995. Population trend estimates from reproductive and survival data. Pages 13–19 in R.R. Knight and
B.M. Blanchard, editors. Yellowstone grizzly bear
investigations: Annual Report of the Interagency Study
Team, 1994. National Biological Service, Bozeman,
Montana, USA.
ELDRIDGE, M.D.B., J.E. KINNEAR, AND M.L. ONUS. 2001.
Source populations of dispersing rock-wallabies (Petrogale lateralis) identified by assignment tests on
multilocus genotypic data. Molecular Ecology
10:2867–2876.
FORTIN, J. 2009. Assessing habitat and diet selection for
grizzly bear and American black bear in Yellowstone
National Park: 2008 annual progress report.
Pages 48–49 in C.C. Schwartz, M.A. Haroldson, and
K. West, editors. Yellowstone grizzly bear investigations: Annual report of the Interagency Grizzly Bear
Study Team, 2008. US Geological Survey, Bozeman,
Montana, USA.
HAROLDSON, M.A., C.C. SCHWARTZ, AND G.C. WHITE. 2006.
Survival of independent grizzly bears in the Greater
Yellowstone Ecosystem, 1983–2001. Pages 33–42 in
C.C. Schwartz, M.A. Haroldson, G.C. White, R.B.
Harris, S. Cherry, D. Moody, and C. Servheen, editors.
Temporal, spatial, and environmental influences on the
demographics of the Yellowstone grizzly bear. Wildlife
Monographs 161.
———. 2008. Assessing trend and estimating population
size from counts of unduplicated females. Pages 9–14
in C.C. Schwartz, M.A. Haroldson, and K. West,
editors. Yellowstone grizzly bear investigations: Annual report of the Interagency Grizzly Bear Study
Team, 2007. US Geological Survey, Bozeman, Montana, USA.
———, K.A. GUNTHER, AND T. WYMAN. 2008a. Nature
note: Possible grizzly cub adoption in Yellowstone
National Park. Yellowstone Science 16(2):42–44.
Ursus 21(1):1–13 (2010)
11
———, C.C. SCHWARTZ, AND K.A. GUNTHER. 2008b.
Grizzly bears in the Greater Yellowstone Ecosystem:
Garbage, controversy, and decline to recovery. Yellowstone Science 16(2):13–24.
HARRIS, R.B., C.C. SCHWARTZ, M.A. HAROLDSON, AND
G.C. WHITE. 2006. Trajectory of the Yellowstone
grizzly bear population under alternative survival
rates. Pages 44–54 in C.C. Schwartz, M.A. Haroldson,
G.C. White, R.B. Harris, S. Cherry, D. Moody, and
C. Servheen, editors. Temporal, spatial, and environmental influences on the demographics of the Yellowstone grizzly bear. Wildlife Monographs 161.
———, G.C. WHITE, C.C. SCHWARTZ, AND M.A. HAROLDSON. 2007. Population growth of Yellowstone
grizzlies: Uncertainty, correlation, and future monitoring. Ursus 18:167–177.
INTERAGENCY GRIZZLY BEAR COMMITTEE. 1986. Interagency
grizzly bear guidelines. US Forest Service, Washington,
DC, USA.
JONKEL, C.J. 1982. Five year summary report. Border
grizzly project special report number 60. University of
Montana, Missoula, Montana, USA.
KENDALL, K.C., J.B. STETZ, D.A. ROON, L.P. WAITS, J.B.
BOULANGER, AND D. PAETKAU. 2008. Grizzly bear
density in Glacier National Park, Montana. Journal
of Wildlife Management 72:1693–1705.
———, ———, J. BOULANGER, A.C. MACLEOD, D.
PAETKAU, AND G.C. WHITE. 2009. Demography and
genetic structure of a recovering brown bear population. Journal of Wildlife Management 73:3–17.
KNIGHT, R., AND B. BLANCHARD. 1983. Report of the
Interagency Study Team, 1982. National Park Service
report, Bozeman, Montana, USA.
———, AND L. EBERHARDT. 1984. Projected future
abundance of the Yellowstone grizzly bear. Journal of
Wildlife Management 48:1434–1438.
———, B. BLANCHARD, AND D.J. MATTSON. 1985. Report
of the Interagency Study Team, 1983–84. National Park
Service, Bozeman, Montana, USA.
———, AND L. EBERHARDT. 1985. Population dynamics of
the Yellowstone grizzly bear. Ecology 66:323–334.
———, AND ———. 1987. Prospects for Yellowstone
grizzly bears. International Conference on Bear Research and Management 7:45–50.
———, B.M. BLANCHARD, AND L.L. EBERHARDT. 1995.
Appraising status of the Yellowstone grizzly bear
population by counting females with cubs-of-the-year.
Wildlife Society Bulletin 23:245–248.
LONGMIRE, J.L., A.K. LEWIS, N.C. BROWN, J.M. BUCKINGHAM, L.M. CLARK, M.D. JONES, L.J. MEINKE, J.M.
MEYNE, R.L. RATLIFF, F.A. RAY, R.P. WAGNER, AND
R.K. MOYZIS. 1988. Isolation and molecular characterization of a highly polymorphic centromeric
tandem repeat in the family Falconidae. Genomics
2:14–24.
12
DNA SUPPORTS ISOLATION OF GYE GRIZZLY N Haroldson et al.
MACe, R.D., AND C.J. JONKEL. 1986. Local food habits of
the grizzly bear in Montana. International Conference
on Bear Research and Management 6:105–110.
MARSTON, R.A., AND J.E. ANDERSON. 1991. Watersheds
and vegetation of the Greater Yellowstone Ecosystem.
Conservation Biology 5:338–346.
MATTSON, D.J., B.M. BLANCHARD, AND R.R. KNIGHT.
1991. Food habits of Yellowstone grizzly bears.
Canadian Journal of Zoology 69:1619–1629.
———, R.G. WRIGHT, K.C. KENDALL, AND C.J. MARTINKA. 1995. Grizzly bears. Pages 103–105 in E.T.
LaRoe, G.S. Farris, C.E. Puckett, P.D. Doran, and
M.J. Mac, editors. Our living resources: A report to the
nation on the distribution, abundance, and health of
U.S. plants, animals, and ecosystems. US Department
of the Interior, National Biological Service, Washington, DC, USA.
MCARDLE, B.H. 1990. When are rare species not there?
Oikos 57:276–277.
MCLELLAN, B.N., AND F.W. HOVEY. 2001. Natal dispersal
of grizzly bears. Canadian Journal of Zoology
79:838–844.
MILLER, C.R., AND L.P. WAITS. 2003. The history of
effective population size and genetic diversity in the
Yellowstone grizzly (Ursus arctos): Implications for
conservation. Proceedings of the National Academy of
Sciences, USA 100:4334–4339.
NATIONAL ACADEMY OF SCIENCES. 1974. Committee on the
Yellowstone grizzlies. Division of Biological Sciences,
Assembly of Life Sciences, National Research Council,
National Academy of Sciences, US Department of the
Interior, Washington DC, USA.
OSTRANDER, E., G. SPRAGUE, AND J. RHINE. 1993.
Identification and characterization of dinucleotide
repeat (CA)n markers for genetic mapping in dog.
Genomics 16:207–213.
PAETKAU, D., L.P. WAITS, P. WASER, L. CLARKSON, L.
CRAIGHEAD, E. VYSE, R. WARD, AND C. STROBECK. 1998.
Variation in genetic diversity across the range of North
American brown bears. Conservation Biology 12:418–429.
———. 2003. An empirical exploration of data quality in
DNA-based population inventories. Molecular Ecology
12:1375–1387.
———, R. SLADE, M. BURDEN, AND A. ESTOUP. 2004.
Genetic methods for the direct, real-time estimation of
migration rate: A simulation-based exploration of
accuracy and power. Molecular Ecology 13:55–65.
PRITCHARD, J.K., M. STEPHENS, AND P. DONNELLY. 2000.
Inference of population structure using multilocus
genotype data. Genetics 155:945–959.
PROCTOR, M., B.N. MCLELLAN, AND C. STROBECK. 2002.
Population fragmentation of grizzly bears in southeastern British Columbia, Canada. Ursus 13:153–160.
———, ———, ———, AND R.M.R. BARCLAY. 2004.
Gender-specific dispersal distances of grizzly bears
estimated by genetic analysis. Canadian Journal of
Zoology 82:1108–1118.
———, ———, ———, AND ———. 2005. Genetic
analysis reveals demographic fragmentation of grizzly
bears yielding vulnerably small populations. Proceedings of the Royal Society, London 272:2409–2416.
———, D. PAETKAU, B.N. MCLELLAN, G. STENHOUSE,
K.C. KENDALL, R.D. MACe, W.F. KASWORM, C.
SERVHEEN, C. LAUSEN, M.L. GIBEAU, W.L. WAKKINEN,
M.A. HAROLDSON, G. MOWAT, L.M. CIARNIELLO, R.
BARCLAY, M.S. BOYCE, C. STROBECK, AND C.C.
SCHWARTZ. In press. Population fragmentation and
inter-ecosystem movements of grizzly bears in western
Canada and the northern USA. Wildlife Monographs.
PYARE, S., S. CAIN, D. MOODY, C. SCHWARTZ, AND J.
BERGER. 2004. Carnivore re-colonisation: Reality,
possibility and a non-equilibrium century for grizzly
bears in the Southern Yellowstone Ecosystem. Animal
Conservation 7:1–7.
ROON, D.A., L.P. WAITS, AND K.C. KENDALL. 2005. A
simulation test of the effectiveness of several methods
for error-checking non-invasive genetic data. Animal
Conservation 8:203–215.
ROUSSET, F. 2008. Genepop ‘007: A complete reimplementation of the Genepop software for Windows and
Linux. Molecular Ecology Resources 8:103–106.
SANDSTROM, P.L. 1996. Identification of potential linkage
zones for grizzly bears in the Swan-Clearwater Valley
using GIS. Thesis, University of Montana, Missoula,
Montana, USA.
SERVHEEN, C. 1999. Status and management of the grizzly
bear in the lower 48 United States. Pages 50–54 in
C. Servheen, S. Herrero, and B. Peyton, compilers.
Bears: Status survey and Conservation Action Plan.
IUCN/SSC Bear and Polar Bear Specialist Groups.
International Union for the Conservation of Nature,
Gland, Switzerland and Cambridge, UK.
———, J.S. WALLER, AND P. SANDSTROM. 2001. Identification and management of linkage zones for grizzly
bears between the large blocks of public land in the
Northern Rocky Mountains. Pages 161–179 in C.L.
Irwin, P. Garrett, and K. McDermott, editors. Proceedings of the International Conference of Ecology
and Transportation. Center for Transportation and the
Environment, North Carolina State University, Raleigh, North Carolina, USA.
SCHWARTZ, C.C., M.A. HAROLDSON, K.A. GUNTHER, AND
D. MOODY. 2002. Distribution of grizzly bears in the
Greater Yellowstone Ecosystem, 1990–2000. Ursus
13:203–212.
———, ———, ———, AND ———. 2006a. Distribution
of grizzly bears in the Greater Yellowstone Ecosystem,
in 2004. Ursus 17:63–66.
———, ———, AND S. CHERRY. 2006b. Study area and
methods for collecting and analyzing demographic data
Ursus 21(1):1–13 (2010)
DNA SUPPORTS ISOLATION OF GYE GRIZZLY N Haroldson et al.
on the Yellowstone grizzly bear. Pages 9–16 in C.C.
Schwartz, M.A. Haroldson, G.C. White, R.B. Harris,
S. Cherry, D. Moody, and C. Servheen. Temporal,
spatial, and environmental influences on the demographics of the Yellowstone grizzly bear. Wildlife
Monographs 161.
STONEBERG, R.P., AND C.J. JONKEL. 1966. Age determination of black bears by cementum layers. Journal of
Wildlife Management 30:411–414.
TABERLET, P., J. CAMARRA, AND S. GRIFFIN. 1997.
Noninvasive genetic tracking of the endangered Pyrenean brown bear population. Molecular Ecology
6:869–876.
TAYLOR, M.K., M. KUK, M. OBBARD, H.D. CLUFF, AND B.
POND. 2001. RISKMAN: Risk analysis for harvested
populations of age structured, birth-pulse species.
Software manual. Ontario Ministry of Natural Resources, Peterborough, Ontario, Canada. Available at
http://riskman.nrdpfc.ca/riskman.htm.
US FISH AND WILDLIFE SERVICE. 1993. Grizzly bear
recovery plan. Missoula, Montana, USA.
Ursus 21(1):1–13 (2010)
13
WAITS, L.P., D. PAETKAU, C. STROBECK, AND R.H. WARD.
1998. A comparison of genetic variability in brown
bears (Ursus arctos) from Alaska, Canada, and the
lower 48 states. Ursus 10:307–314.
WALKER, R., AND L. CRAIGHEAD. 1997. Analyzing wildlife
movement corridors in Montana using GIS. Environmental Sciences Research Institute. Proceedings of the
1997 International ESRI Users conference. Available at
http://proceedings.esri.com/library/userconf/proc97/proc97/
to150/pap116/p116.htm.
WOODS, J.G., D. PAETKAU, D. LEWIS, B.N. MCLELLAN, M.
PROCTOR, AND C. STROBECK. 1999. Genetic tagging of
free-ranging black and brown bears. Wildlife Society
Bulletin 27:616–627.
ZEDROSSER, A., O. STØEN, S. SAEBO, AND J.E. SWENSON.
2007. Should I stay or should I go? Natal dispersal in
the brown bear. Animal Behavior 74:369–376.
Received: 10 August 2009
Accepted: 27 February 2010